How could the skilled mavens at the core of a more successful shared online site for independent music be recruited, screened, and retained?  The simplest way would be to compensate them with a share of the sales, streaming plays, or other revenue-generating events that reflects the actual influence of their recommendations.

Now, in some quarters, this approach might evoke gasps or shudders.  “How can you pay people for recommending things?” some might ask.  “Won’t that induce them to just blindly recommend everything?”

The appropriate answer is, “Not really.”  The mere act of making a recommendation is not what they’d be paid for.  In order for a recommender to make any money, visitors would need to have first trusted that maven enough to choose him as a guide to good songs, and then liked what they found.  Mavens who recommend things indiscriminately wouldn’t be chosen very often.  They’d be like stores that don’t exercise judgment in what they put on their shelves:  nobody wants to shop in such a sprawling mess.  They wouldn’t make a dime, and they’d ruin their “brand” in the process.

Or to approach the question from a different angle, lots of people are already paid to recommend songs—along with movies, books, restaurants, and a variety of other things—without ill effects.  These people are called critics.  They don't make recommendations indiscriminately, because they know that if people stop trusting their recommendations, they'll lose their prestige, their jobs, and their income.  And it will often be these same critics who make recommendations on a shared music site.

Of course, a good shared site shouldn’t limit its recommenders to just people who already have credentials as critics, since a “golden ear” might be found anywhere.  Anybody ought to be allowed to recommend songs.  They should be presented to visitors, though, in order of who’s adding the most value to the site, in terms of recommendations followed to successful conclusions (like a visitor liking a song well enough to buy it), expressed as a percentage of their total recommendations.  Visitors could choose for themselves how far they want to look down this list, in the same way they now choose how far from the most heavily-trafficked commercial areas they want to go to shop for their clothes.

With properly motivated people constantly unearthing the best songs posted on it, a shared website for independent music would no longer have to be just an undifferentiated dumping ground for songs the traditional music industry had passed by, requiring a lot of patience on the part of visitors to dig through all the forgettable material in search of rare “finds.”  The gems would already be out there on the surface, waiting for visitors to scoop them up.

And since they’d be out there in greater numbers than the songs getting traditional radio airplay, and more closely matched to a greater diversity of tastes, the overall encounter would be richer and more rewarding for consumers than traditional methods of marketing music.  It would be a musical equivalent of the kind of retail experience that’s currently enjoyed only by a fortunate few who happen to reside in areas with especially vibrant and diverse shopping districts.

This solution is actually not far from becoming a reality.  Virtually all the key software needed to support it has already been developed, and currently powers a publicly-accessible proof of concept site.  The main work ahead involves letting people know it’s out there, and getting them to start using it.